INDEX

Explanations

No Explanations Found

New Auto-Interp

Top Features by Cosine Similarity

Configuration

ckkissane/attn-saes-gpt2-small-all-layers/gpt2-small_L5_Hcat_z_lr1.20e-03_l11.00e+00_ds49152_bs4096_dc1.00e-06_rsanthropic_rie25000_nr4_v9.pt

Prompts (Dashboard)

36,864 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

49,152

Data Type

float32

Hook Name

blocks.5.attn.hook_z

Hook Layer

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Head Attr Weights

0:0.08

1:0.07

2:0.07

3:0.07

4:0.09

5:0.08

6:0.09

7:0.07

8:0.09

9:0.08

10:0.07

11:0.07

Negative Logits

 Intercept

-1.83

 undercover

-1.61

 capt

-1.60

 Investigations

-1.59

 infring

-1.59

 compr

-1.55

 expose

-1.51

 transl

-1.49

 downloaded

-1.48

 insert

-1.44

POSITIVE LOGITS

ecause

2.39

ebted

2.02

Û

1.89

JUST

1.85

BLIC

1.83

TRUMP

1.81

 Dhabi

1.75

isans

1.67

 Buddh

1.66

metics

1.65

Activations Density 0.000%

No Known Activations

This feature has no known activations.